Move And Renumber The Origin Of Circular Genome

Question

[ivpz]

Hi all,

After doing a de novo assembly of a circular genome and comparing the linear contig to a closely related circular genome, I realise that the origin on the contig is actually located at a position about 30Kbp from the bp 1. I wonder what may be the best way to move that 30 kbp to a correct location, renumber the contig?

Thank you.

Answer 1

[Istvan Albert]

It all depends what type of data you need to alter.

usually one just shifts the coordinates and applies the boundary condition.

[ivpz]

To clarify, I need to "cut and paste" the 30 Kbp 5'end to the 3'end. I was hoping that there may be a more efficient way to do this.

[Istvan Albert]

if all you have is a sequence then just move the sequence around, you can also do that with tools of various kinds if you wanted to use numerical coordinates, like so:

# create the index
$ samtools faidx chrI.fa

# select a region 
$ samtools faidx chrI.fa chrI:1-10
>chrI:1-10
CCACACCACA

# select another region
$ samtools faidx chrI.fa chrI:100-110
>chrI:100-110
GGCCAACCTGT

now you could just place these into a file and concatenate like so:

$ samtools faidx chrI.fa chrI:1-10 | grep -v ">" > a
$ samtools faidx chrI.fa chrI:100-110 > b
$ cat a >> b

by the end of the operation file b will contain the sequences from 100-110 followed by 1-10

[ivpz]

I've done something similar using extractseq from emboss. I was thinking more of a simple one liner :)

Anyway, thanks for the suggestion. I guess samtools may work faster.